30 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS during a 2-week period before antiperspirant treatment was started. Stoughton, Chiu, Fritsch and Nurse (25) utilized $0-min collection periods. The pad method has proved to be extremely useful for the assessment of axillary sweating and despite the possible errors inherent in the technique (weighing errors, evaporative loss from the pads, lack of distinction between sweat and sebum, variation in position of the pad in the axillary vault) it is widely used as a routine technique for testing antiperspirant products. In testing antiperspirants it is usual to take 'control' collections during a period when neither axilla is being treated in the ensuing period the anti- perspirant under test is applied to one axilla only, the other remaining un- treated. Generally, the axilla in which sweating is most profuse receives applications of the test antiperspirant. To interpret the results of such tests, it is necessary to know (and possibly to standardize) the amount of antiperspirant material applied to the skin, either as a total dose or as a dose per unit area of skin, along with the time interval between application of test material and measurement of inhibition. Treatment of results Frequently the weight of sweat produced in each axilla is totalled and the average ratio of one axillary total to the other is determined. The per- centage reduction in sweating caused by application of antiperspirant is calculated: Treatment week ratio 1 X 100 (27) Ratio prior to treatment Wooding (50), however, suggests that in gravimetric tests on axillary sweating the geometric mean, rather than the arithmetic mean, should be used in the calculation in this case all the weights should be converted into the natural logarithms prior to statistical analysis being carried out. INSTRUMENTAL TECHNIQUES In the last 20 years precision instrumental techniques have been developed for quant, flying sweat gland act!v!t2. These ,metho4s'mostly involve the use of a collection chambe• attached'to thee s, ki, n su•fac• ,through which a stream of dry gas, or gas with a known low relative humidity

MEASUREMENT AND CONTROL OF PERSPIRATION 31 is passed. The moisture picked up by the gas as it passes through the col- lection chamber represents the moisture lost from the enclosed skin area. The gas used is generally air, oxygen, or nitrogen and the moisture collected by this stream of gas can be measured in one of the following ways: (i) gravimetrically, following condensation in a collection coil (8, 27, 29, 51- 54). (ii) by humidity-sensing elements which undergo conductivity or resistance changes with variations in the RH of air passing over them (17, 52, 53, 56-58). (iii) by absorption of infra-red radiation (59, 60). Collection coils Kuno (8) used a celluloid collection chamber through which gas dried by passing over calcium chloride or phosphoric arthydride was passed. Mois- ture picked up by the gas stream was trapped in fared U-tubes filled with calcium chloride. In place of such U-tubes, Pinson (51) used flasks con- taining pumice stone saturated with concentrated sulphuric acid. Since cellulose acetate allows slow diffusion of water Neumann, Cohn and Burch (52) used a brass collecting chamber and aluminium coils im- mersed in an alcohol-dry ice freezing mixture as this was considered more effective in removing the moisture than calcium chloride. The authors claimed that this method had an accuracy of 2.6%. Jenkins (27) published a similar method, but collections were carried out in the axillae where metal cups were held in position with rubber straps. An air flow of 1.8-2.2 1 min- 1 was considered adequate for complete removal of moisture from each axilla, whilst still allowing complete con- densation in the coils. Often in these tests the mass of sweat collected is only a fraction of the mass of the coil, and great care must be taken in measurement. The coils must be thoroughly dry inside and out when their mass is first determined and before the subsequent determination is made they must be dried on the outside and any air in them brought to atmospheric pressure. Humidity sensing elements These elements generally consist of a small plate coated with a thin film of a material whQse e!ectrical conductivity (or resistance) varies in pro- portion to the moisture-content of a stream 9f air passing over it. When used in perspiration studies, air from a collection chamber attached to the